Class VariableEliminationInference

java.lang.Object

edu.cmu.tetrad.bayes.VariableEliminationInference

All Implemented Interfaces:

Serializable

public final class VariableEliminationInference
extends Object
implements Serializable

Exact inference engine over a BayesIm supporting:

• Soft evidence via allowed categories (Proposition)
• Hard evidence via node=category
• Marginals, conditionals, and joint marginals (all conditional on evidence)

This class is designed specifically to satisfy the methods required by JunctionTreeUpdater. 2026-02-21 jdramsey + (VE-based implementation provided here)

See Also:

• Serialized Form

Constructor Summary

Constructors

Constructor	Description
VariableEliminationInference(BayesIm bayesIm)	Construct inference over the given BayesIm.

Method Summary

Modifier and Type	Method	Description
double[]	getConditional(int node, int[] parents, int[] parentValues)	Conditional distribution P(node | parents=parentValues, evidence).
double	getJointProbability(int[] vars, int[] values)	Conditional joint marginal P(vars=values | evidence).
double	getMarginal(int node, int category)	Conditional marginal P(node=category | evidence).
void	setAllowedCategories(Proposition allowedCategories)	Soft evidence: set allowed categories for each variable.
void	setEvidence(int node, int category)	Hard evidence: set node = category.

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

VariableEliminationInference

public VariableEliminationInference(BayesIm bayesIm)

Construct inference over the given BayesIm.

Parameters:

bayesIm - BayesIm to perform inference over

Method Details

setAllowedCategories

public void setAllowedCategories(Proposition allowedCategories)

Soft evidence: set allowed categories for each variable. Proposition must be indexed to this BayesIm.

Parameters:

allowedCategories - Proposition specifying allowed categories for each variable

setEvidence

public void setEvidence(int node,
 int category)

Hard evidence: set node = category. Pass a negative category to clear hard evidence for that node.

Parameters:

node - Node

category - Category

getMarginal

public double getMarginal(int node,
 int category)

Conditional marginal P(node=category | evidence).

Parameters:

node - Node

category - Category

Returns:

Conditional marginal probability

getConditional

public double[] getConditional(int node,
 int[] parents,
 int[] parentValues)

Conditional distribution P(node | parents=parentValues, evidence). This returns an array of length numCategories(node).

Parameters:

node - Node

parents - Parent nodes

parentValues - Values for parent nodes

Returns:

Conditional distribution

getJointProbability

public double getJointProbability(int[] vars,
 int[] values)

Conditional joint marginal P(vars=values | evidence). (This is what callers typically mean by "joint marginal".)

Parameters:

vars - Variables

values - Values for variables

Returns:

Joint probability